Switches, Hubs, and Modems
cancel
Showing results for 
Search instead for 
Did you mean: 

Server-to-switch distributed trunking or VRRP?

groque
Frequent Advisor

Server-to-switch distributed trunking or VRRP?

Hi forum,

I am looking to buy two 5400zl's for my distribution and core switches.

I will have my SAN, VM servers and Access switches connecting to the core

Should I purchase the VRRP license for redundancy or should I use the server-to-switch distributed trunking feature?

Is it possible to have a 2Gb trunk between both 5400yl's and have them do server-to-switch distributed trunking and connect my SAN, VM and Access switches to both of them

Or

Use VRRP in an active/active scenario?

Thanks for the reposes.

3 REPLIES
Ralf Krause
Frequent Advisor

Re: Server-to-switch distributed trunking or VRRP?

Hi grogue,
the fact that you at least think about VRRP, tells me that the switches are going to act as default gateway for your nodes.

If so, this will require "ip routing" enabled on the switches.

IP routing and distributed trunking are mutually exclusive (not possible to be configured both at a time on the same box!).

So, take care here!

I guess this leads to VRRP licensing.

Hope that helps,
Ralf
ISoliman
Super Advisor

Re: Server-to-switch distributed trunking or VRRP?

This limitation has been removed, below updated details regarding DT:

 

http://h20195.www2.hp.com/v2/GetPDF.aspx/4AA3-4841ENW.pdf

 

Even if you are going to use DT you will need to use VRRP aswell, DT is for link aggregation and clients still need gateway so if 1 switch went down the other should take the ROUTING and gateway role.

 

Tony0001
Occasional Visitor

Re: Server-to-switch distributed trunking or VRRP?

By anychance do you know how distributed trunking behaves when a switch fails..?

 

I have Sw1 connected to Sw2 and also have VRRP between them, when one switch fails failover works fine i only lose 1 ping but when the failed switch is back online i lose 4-6 pings and everything is back to normal then.

 

If i remove Distributed trunking and just use Vrrp all works fine.